We have focused in three related areas: 1. the study of vFLIP, a KSHV latent gene product expressed in KSHV-infected cell targets and in Kaposi's sarcoma (KS), Primary Effusion Lymphoma (PEL) and Multicentric Castleman's disease (MCD);2. the study of vIL-6, a KSHV viral product with structural similarity to cellular IL-6, and its role in the athogenesis of MCD;3. Peculiar clinical presentations of MCD;and 4. the development of new therapies for KSHV-induced malignancies occurring in AIDS patients. One of the characteristic features of KSHV is its ability to infect endothelial cells,and to indirectly promote angiogenesis and lymphangiogenesis predominantly by promoting the recruitment of cells that produce pro-angiogenic factors and promoting the expression of pro-angiogenic genes by the cells it infects. ORFK13/vFLIP encodes a 188-amino acid protein, which binds to the Ikb kinase (IKK) complex to activate NFkB. We examined ORFK13/vFLIP contribution to KS phenotype and potential for therapeutic targeting. To this end, we have retrovirally transduced ORFK13/vFLIP into primary human endothelial cells and examined the contribution of this gene to KS phenotype. We found that ORFK13/vFLIP induces the spindle morphology distinctive of KS cells and promotes formation of abnormal vascular networks typical of the disorderly KS vasculature. Microarray analysis of gene expression in endothelial cells transduced with ORFK13/vFLIP detected increased expression of certain pro-inflammatory cytokines, chemokines, and interferon-responsive genes. This study represents the first comprehensive analysis of gene regulation by KSHV-vFLIP. As one might expect from stimulation of pro-inflammatory cytokines and chemokines, we found that ORFK13/vFLIP stimulates adhesion of inflammatory cells characteristic of KS lesions. In additional experiments, we found that that KSHV K13 induces the expression of the NF-kB regulatory proteins A20, ABIN-1 and ABIN-3 in primary human microvascular endothelial cells, and that KS spindle cells express A20 in KS tissue. In reporter assays, A20 strongly impaired K13-induced NF-kB activation in 293T cells, but ABIN-1 and ABIN-3 did not. Mutational analysis established that the C-terminal domain (residues 427-790) is critical for A20 PEL is a fatal viral malignancy in humans, which typically presents as a malignant effusion that later disseminates. In spite of therapy with high-dose chemotherapy or other therapies, PEL is a rapidly fatal malignancy. Rapamycin, which targets mTOR (mammalian target of rapamycin), an effector of cell signaling pathways often deregulated in cancer, showed efficacy against a variety of tumors, particularly thomodulation of NF-kB. In functional assays, A20 inhibited K13-induced secretion of IP-10, and reduced K13-induced cell proliferation. Thus, we demonstrate that A20 negatively regulates NF-kB activation directly induced by KSHV K13. By attenuating excessive and prolonged NF-kB activation induced by K13 that could be harmful to KSHV-infected cells, A20 likely plays an important role in the pathogenesis of KSHV-associated diseases, in which K13 is expressed. Previous studies from our group characterized vIL-6 as an early lytic gene expressed by KSHV, homologous to the cellular IL-6 cytokine. Unlike cellular IL-6, we found that vIL-6 can directly bind and signal through the gp130/JAK-STAT pathway without a requirement for the cellular IL-6 receptor. gp130 is a fairly ubiquitous transmembrane protein;thus, in contract to the cellular cytokine, vIL-6 can activate virtually all cells in the body. Previously, we have transduced vIL6 in NIH3T3 cells;when these cells were transferred to T-cell immunodeficient mice, they generated tumors at a significantly higher rate than control NIH3T3 cells and, importantly, the mice developed splenomegaly, hepatomegaly and plasmacytosis in many tissues. These features are common to patients with MCD. We have now produced transgenic mouse lines in which vIL-6 is ubiquitously expressed. These mice were found to exhibit vIL-6 serum levels comparable with those observed in KSHV-infected patients, to contain elevated amounts of phosphorylated STAT3 in spleen and lymph nodes, where vIL-6 was abundantly produced, and to spontaneously develop key features of human plasma cell-type MCD, including splenomegaly, multifocal lymphadenopathy, hypergammaglobulin-emia, and plasmacytosis. Interestingly, expression of the vIL-6 transgene in cellular IL-6-deficient mice did not yield the MCD-like phenotype observed in IL-6-competent mice. This indicated that endogenous cellular IL-6 is a critical co-factor in the natural history of MCD. We have previously reported that vIL-6 can induce the expression of cellular IL-6 in vitro, and that vIL-6 and cellular IL-6 are often detected at abnormally high levels in patients with MCD. These observations suggest that human IL-6 plays an important role in the pathogenesis of KSHV-associated MCD. Thus, in collaboration with the HAMB group at NCI we have initiated a clinical trial exploring the utility of blocking cellular IL-6 in the treatment of MCD. MCD is characterized by inflammatory symptoms, lymphadenopathy, splenomegaly and cytopenia. KSHV is the cause of virtually all cases of MCD occurring in HIV-infected patients. MCD lesions characteristically contain HHV-8-infected polyclonal IgMl plasmablasts. We have identified three patients who presented with severe symptoms of MCD and had a rapid expansion of circulating B-cells (44-81%) attributable to HHV-8 positive cells sharing the phenotype (IgMl, CD19+, CD20-, CD138-) of HHV-8-infected cells from MCD lesions. These patients displayed a very high HHV-8 viral load in blood (7 logs HHV-8 DNA copies/ml) and high levels of serum vIL-6, the viral homolog of human interleukin 6. Serum IL-6 and IL-10 were also abnormally elevated. Thus, we have expanded the spectrum of HHV-8-related plasmablastic lymphoproliferative disorders in HIV-infected patients to include HHV-8+ polyclonal IgMl B-cell lymphocytosis. At onset, this lymphoproliferative disorder may mimick plasmablastic leukemia/lymphoma. In other experiments we have examined the biochemical basis for diversity of phenotype within KS cells that are KSHV-infected. Such diversity has created uncertainties on the origin of KS tumor cells. We have examined the possibility that KSHV infects endothelial cells and turns them into mesenchymal cells. Endothelial to mesenchymal transition (EndMT), the process by which endothelial cells convert into mesenchymal cells, plays critical roles during development of the heart, and underlies certain forms of pathological organ fibrosis and tissue ossification. We found that KSHV is an inducer of EndMT. Upon KSHV infection, primary dermal microvascular endothelial cells lose expression of endothelial markers, acquire expression of mesenchymal markers, display new invasive and migratory properties, and exhibit increased survival. We discovered that the canonical Notch signaling pathway and the Notch-induced transcription factors Slug and ZEB1 are deployed by KSHV to induce activation of EndMT, whereas the TGF-beta signaling pathway previously linked to EndMT, is not utilized. The KSHV-infected spindle cells within KS lesions display a complex phenotype with features of endothelial and mesenchymal cells, display evidence of Notch activity and express nuclear ZEB1, features compatible with KSHV-induced EndMT in vivo. These results show that KSHV utilizes the EndMT program to endow endothelial cells with invasiveness and resistance to death, traits that likely contribute to KS progression and KSHV persistence. Targeting Notch signaling emerges as a novel experimental approach to the treatment of KS.
Showing the most recent 10 out of 18 publications
